GPT Engineer Review: The Original AI Coding Agent, Still Relevant?

Review of GPT Engineer

★ 3.5/5 · Updated 2026-06-16

|

What is GPT Engineer?

GPT Engineer is the open-source AI coding agent that started the 'specs to code' trend. Created by Anton Osika in 2023, it now has 70K+ GitHub stars and a community of contributors. The current version supports multiple LLM backends and complex multi-step workflows.

What we like

It's open source. You can self-host, modify, and inspect every line of code. No vendor lock-in, no surprise pricing changes.

The 'spec' workflow is great. You write a clear specification, the AI asks clarifying questions, and then builds. The iterative clarification is a unique feature.

Multi-LLM support. You can use GPT-5, Claude, or local models like Llama 4. Switch providers with a config change.

Active community. The Discord is helpful, the GitHub issues are answered, and the project is actively maintained.

What we don't like

Setup is harder than competitors. You need Python, pip, and API keys. No web interface. No one-click deploy.

No IDE integration. It's a CLI tool. You can't use it inside VS Code or Cursor.

Output quality is variable. The same prompt can produce vastly different results depending on the LLM you pick. You need to know what you're doing.

Smaller ecosystem. No marketplace, no templates, no community plugins. Just the core tool.

Pricing

Free and open source. You pay only for LLM API calls (OpenAI, Anthropic, or your own model).

Who is it for?

Devs who want an open-source AI coding agent and don't mind CLI tools. Hobbyists experimenting with different LLMs.

Verdict

★ 3.5/5. Still relevant for open-source fans, but the commercial tools (Cursor, Bolt) are easier and more polished.

|

Visit GPT Engineer →

← Back to all reviews

Related on saas.pet